[Ocfs2-devel] [PATCH] ocfs2: fix __ocfs2_cluster_lock() dead lock

[Wengang Wang]

Hi Joel,

On 10-01-06 18:00, Joel Becker wrote:
> On Wed, Jan 06, 2010 at 04:34:44PM +0800, Wengang Wang wrote:
> > there is deadlock possibility in __ocfs2_cluster_lock().
> > the case is like following.
> > 
> > #in time order
> > 1) node A(the lock owner) is doing an up-convert(UC, PR->EX). it got the lock
> > but didn't check BUSY flag again.
> > 
> > 2) node B requested an UC on the same lock resource. since node A has an EX on
> > the lockres, a bast is issued and OCFS2_LOCK_BLOCKED flag is set to the lockres
> > meanting that a DC is should be done. the DC asks for agreement of node A to
> > release EX. and for now node A refuses that(still in process of getting that
> > lock:) ).
> > 
> > 3) the UC on node A runs into the phrase of checking OCFS2_LOCK_BLOCKED flag and
> > if it can continue without waiting for any DC. code is:
> > 
> >         if (lockres->l_flags & OCFS2_LOCK_BLOCKED &&
> >             !ocfs2_may_continue_on_blocked_lock(lockres, level)) {
> > 
> > analysis:
> > the BLOCKED flag is set in 2), and the UC can't continue to get the lock(
> > ocfs2_may_continue_on_blocked_lock() returns false), so it the DC to finish.
> > 
> > so both UC and DC are waiting for each other --a dead lock.
> 
> 1) Do you have a test case?  Have you seen this happen, or are you just
>    postulating from reading the code?

yes, I have test case for this. when I was testing freeze/thaw, it was
hit. while, the code for freeze/thaw is not sent out yet :P

the main cluster lock iterations is like this:
0) a cluster lock for freeze/thaw is added. it works like open lock on
inode. so node A and node B both have PR lock on it.

1) node A want to freeze the fs, so requested for a EX.

2) node B released PR. and very immediately requests a PR waiting for
thawing.

3) things described above happened.

regards,
wengang.
> 2) It sure looks like you're right.
> 3) I hate our locking spaghetti!